Abstract
Construction robots, as a disruptive innovation in the construction industry, can effectively resolve industry pain points such as low construction efficiency, high safety risks, and high labor costs. For spraying robots of building exterior wall, the adoption of Human-Computer Interaction (HCI) control technology can effectively improve spraying quality, save raw materials, reduce construction costs, and enhance construction efficiency while ensuring personnel safety. The authors developed a prototype of a spraying robot based on HCI technology and proposed a real-time control method based on dynamic gesture recognition. This method focuses on the distance changes of 21 joints on the hand, combines the Support Vector Machine (SVM) classification algorithm, and constructs a real-time closed-loop control system including gesture acquisition, algorithm analysis, and robot execution. By capturing the movement trajectory of the right index finger and recognizing gestures of the left hand, the maximum error of turning head is only 0.89°, and the maximum error of the telescopic rod is only 0.21 mm. Through comparative experiments with automatic trajectory operation mode and HCI control mode at construction site, the HCI control mode can still maintain good spraying effects under different wind speed interference conditions, ensuring the continuity and accuracy of spraying quality.